The Unionist Pact: Did it work?
Now that the dust has settled following #GE2015, Dr Christopher Raymond examines the 'Unionist Pact' in Northern Ireland and asks whether it was successful.
The 2015 Westminster election saw the Democratic Unionist Party (DUP) and the Ulster Unionist Party (UUP) in Northern Ireland agree a pact in order to maximise the representation of unionists at Westminster. In this pact, the UUP agreed not to contest Belfast East and Belfast North, while the DUP agreed not to contest Fermanagh & South Tyrone and Newry & Armagh. Because the first-past-the-post system rewards the party with the largest vote share (regardless how small that vote share is), the division of voters between the DUP and UUP might hurt both parties’ shared interest in seeing unionist representation at Westminster. By combining their vote shares, the party leaders hoped to prevent non-unionist parties from winning at the expense of two smaller parties dividing the unionist vote. This article analyses the effectiveness and limits of the unionist pact.
Three seats won
The biggest claim by supporters in favour of the pact’s effectiveness was that three out of four of the targeted seats were won by the agreed unionist candidate. While the DUP held onto its seat in North Belfast, it regained Belfast East and the UUP won Fermanagh & South Tyrone. The unionist pact only failed to produce victory in Newry & Armagh.
The logic underlying this argument can be seen in the case of Belfast East. In 2010, the DUP and UUP won fairly sizable vote shares: 32.8 and 21.2 percent, respectively. Had the two parties fielded only one candidate in 2010, they would have commanded a majority of the vote (54.0 percent). Instead, because they divided the unionist vote, the Alliance Party was able to win the seat with a mere 37.2 percent of the vote. By cooperating in 2015, the DUP won 49.3 percent of the vote, edging out the Alliance Party’s 42.8 percent and taking back the seat. A similar logic motivating the pact can be seen in the other three seats as well.
The pact and unionist voters
One concern with the pact was that some unionist voters might find the pact distasteful and either cast a protest vote against the agreed candidate or stay home on election day. In fact, however, there is little evidence the pact hurt the agreed candidates’ vote shares. Table 1 presents the vote shares for the agreed unionist candidate in each constituency, as well as the estimated vote shares for each party’s candidate assuming the pact had never been agreed.[1]
Table 1: The Estimated Benefits of the Pact
Entries in the first row are the results for the agreed unionist candidate in 2015; entries in the second and third rows are the estimated vote shares for the individual DUP and UUP candidates, respectively, while entries in the bottom row are the surplus vote shares due to the unionist pact (which is calculated by subtracting each party’s estimated vote share from the agreed unionist vote share).
In three of the four targeted seats, the pact was actually associated with higher vote shares for the agreed unionist candidate than would have been expected had the two parties competed separately. This estimated benefit ranged from an increase of 5.2 percentage points in Newry & Armagh to an increase of 13.5 percentage points in Fermanagh & South Tyrone. Only Belfast North saw lower shares for the agreed unionist candidate than would have been seen without the pact. Thus, the evidence suggests the pact did not turn off unionist voters, but instead may have actually invigorated them.
Was the pact a success?
While the pact does not appear to have upset many unionist voters, it may be a stretch to call the pact a complete success. Although it appears to have been successful in uniting unionists to defeat non-unionist candidates in Belfast East and Fermanagh & South Tyrone, the pact failed to make Newry & Armagh competitive. Moreover, the pact appears to have been unnecessary in Belfast North: the estimated DUP vote share (42.4 percent) alone would have been sufficient to beat Sinn Féin (at 33.9 percent).
More problematic is the fact that the pact’s ultimate goal — to give Northern Ireland’s unionists greater say at Westminster — may have fallen short. While the pact increased representation of unionist parties by one seat, the pact was motivated in part by the assumption that the two unionist parties would play a role in forming the government after 7 May. The fact the Conservatives won an outright majority meant DUP and UUP MPs have been rendered superfluous. Given their very brief audience with the Prime Minister prior to the formation of the cabinet, it looks like the greater representation of unionists at Westminster has not translated into greater influence.
Quo vadis?
Going forward, it will be interesting to see how the DUP and UUP respond in future elections. While the pact suggests cooperation might yield potential gains for both parties without angering unionist voters, the fact the parties cater to different groups within unionism makes cooperation difficult (as evidenced by the inability to include more than four constituencies in the pact). Moreover, the fact that increased representation failed to produce significant influence at Westminster only weakens incentives for future cooperation.
[1] These estimates are generated by comparing the results from the 2015 election to the average DUP/UUP result in each constituency from the 2001-2010 Westminster elections. I predict each party’s vote shares at the constituency level using information regarding the percentage of Protestants, the percentage of people claiming some sort of benefit, and whether the party was the incumbent in the constituency. For DUP vote shares, I also include information regarding the percentage of degree-holders in each constituency, assuming the impact of this depends on the percentage of Protestants in each constituency.
Dr Christopher Raymond is a Lecturer in Politics at Queen’s University Belfast.
The featured image is published under a Creative Commons license and can only be re-published under the conditions of the license.
